pax-2controls multiple steps of urogenital development · the urogenital system is derived...

4057Development 121, 4057-4065 (1995)Printed in Great Britain © The Company of Biologists Limited 1995DEV2038

Pax-2 controls multiple steps of urogenital development

Miguel Torres1, Emilia Gómez-Pardo1, Gregory R. Dressler2 and Peter Gruss1,*1Abteilung Molekulare Zellbiologie, Max-Planck-Institut für Biophysikalische Chemie, Göttingen 37077, Germany2Howard Hughes Medical Institute and Department of Pathology, University of Michigan, Ann Arbor, MI 48109, USA

*Author for correspondence

Urogenital system development in mammals requires thecoordinated differentiation of two distinct tissues, theductal epithelium and the nephrogenic mesenchyme, bothderived from the intermediate mesoderm of the earlyembryo. The former give rise to the genital tracts, uretersand kidney collecting duct system, whereas mesenchymalcomponents undergo epithelial transformation to formnephrons in both the mesonephric (embryonic) andmetanephric (definitive) kidney. Pax-2 is a transcriptionalregulator of the paired-box family and is widely expressedduring the development of both ductal and mesenchymalcomponents of the urogenital system. We report here thatPax-2 homozygous mutant newborn mice lack kidneys,ureters and genital tracts. We attribute these defects to dys-genesis of both ductal and mesenchymal components of thedeveloping urogenital system. The Wolffian and Müllerian

ducts, precursors of male and female genital tracts, respec-tively, develop only partially and degenerate duringembryogenesis. The ureters, inducers of the metanephrosare absent and therefore kidney development does not takeplace. Mesenchyme of the nephrogenic cord fails toundergo epithelial transformation and is not able to formtubules in the mesonephros. In addition, we show that theexpression of specific markers for each of these componentsis de-regulated in Pax-2 mutants. These data show thatPax-2 is required for multiple steps during the differen-tiation of intermediate mesoderm. In addition, Pax-2mouse mutants provide an animal model for human hered-itary kidney diseases.

Key words: Pax-2, urogenital development, gene targeting, mouse

SUMMARY

INTRODUCTION

Pax genes encode a family of transcriptional regulators specif-ically expressed during the development of a wide range ofstructures and organs (Gruss and Walther, 1992). As indicatedby both human and mouse mutations, at least five Pax geneshave critical morphogenetic functions during the developmentof complex tissues (Chalepakis et al., 1992; Strachan and Read,1994). These functions include the development of paraxialmesoderm derivatives (Pax-1) (Balling et al., 1988), neuralcrest cells and pre-myoblast migration (Pax-3) (Epstein et al.,1991; Bober et al., 1994), B-cell and midbrain development(Pax-5) (Urbanek et al., 1994), and eye and nose development(Pax-6) (Hill et al., 1991; Hadler et al., 1995). A remarkablecharacteristic is the semidominant character of loss-of-functionmutations for three of the members of the family, Pax-1, Pax-3 and Pax-6. This feature suggests that the dosage of Paxprotein is essential to ensure wild-type function.

The urogenital system is derived predominantly from theintermediate mesoderm of the early embryo. This mesodermlies between paraxial and lateral mesoderms and undergoesextensive epithelial transformation to generate the ducts andtubules that compose urogenital tracts and kidneys (see Saxén,1987, for review). The first event in the differentiation of inter-mediate mesoderm is the formation of the Wolffian duct,

which progresses rostrocaudally from the cervical leveltowards the cloaca. The duct runs parallel to a tract of inter-mediate mesoderm, the nephrogenic cord. Subsequently, threeembryonic kidneys, pronephros, mesonephros andmetanephros, develop in spatial and temporal sequence. Firstand most anterior, the pronephros forms by an epithelial trans-formation of the nephrogenic cord mesenchyme and results ina linear array of tubules emptying into the Wolffian duct.While the pronephros seems to represent a true excretory organin fish and amphibians, it remains a rudimentary and transitorystructure in the mouse. The mesonephros appears after and justposterior to pronephros and develops in a very similar manner.However, mesonephric tubules may be functional during theembryonic life of mammals. Metanephric development startswhen the ureter buds out from the posterior Wolffian duct andcontacts the metanephric mesenchyme in the caudal part of thenephrogenic cord. Signals from the ureter induce metanephricmesenchyme to condense and proliferate at the ureter tips andreciprocal signals from the mesenchyme induce the ureter togrow and branch, thus forming the kidney collecting system(Grobstein, 1953, 1955). The induced mesenchyme at the tipsof the branches undergoes transformation to generate glomeru-lar, proximal tubular and distal tubular epithelium.

Mutations in genes expressed during urogenital develop-ment disrupt different aspects of metanephros development.

4058 M. Torres and others

The transcription factor WT-1 is expressed in uninduced mes-enchyme and is required for the induction response (Kreidberget al., 1993). The tyrosine kinase receptor c-ret (Schuchardt etal., 1994) is essential for ureter growth and branching. Thesecreted factor Wnt-4 is required for metanephric tubuleformation (Stark et al., 1994). However, besides a reduction inthe number of mesonephric tubules reported in WT-1 mutants,other epithelial derivatives of intermediate mesoderm developnormally in these mutants.

In the male, mesonephric tubules and the Wolffian duct losetheir excretory function and are transformed into the genitaltract. Genital tracts originate in females from the Müllerianducts, which are also derived from the intermediate mesodermand develop parallel to the Wolffian duct much later in devel-opment. Soon after ducts appear, sexual dimorphism is evidentand the Wolffian duct and mesonephros degenerate in thefemale while the Müllerian duct degenerates in the male (Jostand Magre, 1984).

Pax-2 is expressed in both ductal and mesenchymal com-ponents deriving from the intermediate mesoderm during thedevelopment of pronephros, mesonephros and metanephros(Dressler et al., 1990). In vitro experiments have suggested arole for Pax-2 in the conversion of mesenchyme to epitheliumduring metanephros development (Rothenpieler and Dressler,1993). Recently, kidney hypoplasia has been reported associ-ated with heterozygosity for a human PAX-2 point mutation(Sanyanusin et al., 1995) and a mouse chromosome 19 deletion(Krd) spanning about 400 genes, which includes the Pax-2gene (Keller et al., 1994). These data suggest a role for Pax-2in kidney organogenesis and adds Pax-2 to the class of hap-loinsufficient Pax genes.

Fig. 1. Generation of Pax-2 mutant mice. (A) The targeting construct ancharacteristic for wild-type and mutant alleles. Homologous recombinatnucleotides before the ATG, and the PvuII site inside the paired box-encBglII; WT, wild type; Mut, mutant. (B) Comparative Southern Blot analwild-type and expected mutant bands in clone 132. (C) Results obtainedMendelian segregation of the wild-type and mutant allele is observed at

We report here the generation and analysis of a mouse Pax-2 null mutation (Pax-2−). Our results indicate that Pax-2 isessential for multiple steps during intermediate mesodermdevelopment including the differentiation of both ductal andmesenchymal components of mesonephros and metanephrosand their derivatives. All mutations reported to date affectingurogenital development exclusively disrupt metanephrosdevelopment with only one of the two components involved,ureter or mesenchyme. These data show that Pax-2 is a primaryregulator of urogenital development. We suggest that Pax-2may control a regulatory hierarchy of genes involved in thedifferentiation of intermediate mesoderm.

MATERIALS AND METHODS

Targeting Pax-2 locusA positive-negative selection construct was designed. R1 ES cellswere electroporated and selected as described (Wurst and Joyner,1993) 234 independent resistant clones were screened by genomicsouthern blot using probe ‘a’ (Fig. 1) and BamHI digestion. One clone(no. 132) gave the expected recombinant band. Clone 132 was usedto generate chimeras by morula aggregation (Nagy and Rossant, 1993)and germline transmission of the Pax-2 mutation. F1 and F2 genera-tions were genotyped by genomic southern blot using either probe ‘a’and BamHI digestion or probe ‘b’ and HindIII digestion.

Embryo dissectionsFetuses were collected in PBS and yolk sacs or tails were removedfor DNA analysis by Southern blot. Urogenital systems were dissectedout and cleaned from main blood vessels.

HistologyE12.5 embryos were dissected out in PBS and embryos were fixed in

d expected recombination event with sizes of restriction enzyme digestsion would remove about 4 kb between the NotI site in exon one, 20oding region in exon two. B, BamHI; N, NotI; H, HindIII; P, PvuII; Bg,ysis shows the wild-type bands in the parental ES cell line and both from an F2 litter genotyped with probe ‘b’ and HindIII digestion.birth.

4059�Pax-2 required for urogenital development

Fig. 2. Analysis of the urogenitalsystem defects in Pax-2 mutantmice at E17.5 (A-D) Wholemounts of dissected completeurogenital systems at E17.5 inmale (A,C) and female (B,D)homozygous and heterozygousmutants and wild-type fetus. Notethe reduced size of kidneys inheterozygous female shown in B.(E,F) Detail of the distal region ofmale urogenital tracts.(G,H) Detail of gonads andproximal part of genital tracts infemale and male, mutant andwild-type fetus. Note the absenceof seminal vesicles and vasdeferens in F and the normalappearance of gonads in bothsexes (G,H). Note that endoderm-derived components: bladder,urethra and prostatic glands (notshown) are unaffected. Mutantgonads are surrounded by aremnant of the urogenital ridge(*) which is blunt at both ends.ad, adrenal glands; ur, ureter; u,urethra; b, bladder; t, testis; e,epididymis; ut, uterus; o, ovary;od, oviduct; mt, mesonephrictubules developing into ductuliefferentes; v, vas deferens; sv,seminal vesicle.


Fig. 4. Histological analysis at E12.5. (A,C) Cross sections throughgenital ridges of E12.5 control embryos at, respectively, mesonephricand infra-mesonephric levels (B,D) Similar sections from mutantembryos. Black arrowhead shows the place of the degenerating (B)or missing (D) Wolffian duct in mutants, and white arrowhead theplace of the missing mesonephric tubules in mutants. wd, Wolffianduct; mt, mesonephric tubules.

Fig. 3. Analysis of the expression of c-ret during the development of the Wolffian duct inPax-2 mutant embryos. (A,B) Vibratome cross sections of E9.5 embryos after whole-mount in situ hybridization with the c-ret probe; (C,D) similar but correspond to E10.5.(E) Dissected posterior regions of whole-mount E11 wild-type and mutant embryoshybridized with a c-ret probe. wd white arrowhead, Wolffian duct; ub, ureter bud; blackarrow, nephrogenic cord.

4% paraformaldehyde (PFA) and, after dehydration, embedded inparaffin and sectioned at 8 to 10 µm. Sections were stained withhaematoxylin and Eosin. E16.5 embryos were treated similarly exceptthey were fixed in Bouin’s fixative and stained by Mallory’s tetra-chromic procedure.

Antibody stainingE13.5 embryos were fixed overnight in PFA, immersed in sucrose30% for 24 hours, embedded in tissue tek and cryostat sectioned at10 µm. Sections were incubated with a Pax-2 polyclonal antibody asdescribed (Dressler and Douglas, 1992) using a Cy3-conjugatedsecondary antibody.

Whole-mount in situ hybridizations were performed as described(Wilkinson, 1992) using probes previously described for Pax-2(Dressler et al., 1990) and Six-2 (Oliver et al., 1995), and a 700 bpEcoRI fragment from the pmcret10 plasmid of c-ret (Pachnis et al.,1993). After in situ hybridization embryos were embedded in gelatinand vibratome sectioned at 30 µm.

RESULTS

Pax-2 mutants lack kidneys, ureters and genitaltractsIn order to study Pax-2 function, we have generated a nullallele by homologous recombination in embryonic stem (ES)cells (Fig. 1). Pax-2 heterozygous (Pax-2+/−) and homozygous(Pax-2−/−) mutant animals are born at the expected Mendelianproportions. Homozygous mutant animals invariably lackboth ureters and kidneys while heterozygous mutants fre-quently show a reduction in kidney size, which ranges from1/10 of the normal size to wild-type size (Fig. 2). In addition,homozygous mutants of both sexes completely lack the entiregenital tracts; females lack oviducts, uterus and vagina andmales lack ductuli efferentes, epididymis, vas deferens andseminal vesicles. Gonads appear, in mutants from both sexes,surrounded by a blunt-ended remnant of the genital ridge.Kidney defects seem to be the only trait affected in the uro-genital system of heterozygous mutants, genital tracts appearnormal in these embryos. All the structures missing in the

mutants are intermediate mesoderm-derived. In contrast, theendoderm-derived structures: urethra, bladder and prostaticglands appear normal.

Development of genital tracts in Pax-2 mutantsIn males, the epididymis, vas deferens and seminal vesicles arederived from the Wolffian duct and in females, the oviducts,uteri and upper part of the vagina are derived from the Müllerianduct. To study the primary cause of the defects observed, we


Fig. 5. Histological analysis of wolffianand Müllerian ducts development atE17.0. Cross sections of female wild-type (A), female mutant (B), male wild-type (C) and male mutant embryos (D).Note in B the degeneration of theWolffian (*) and Müllerian (arrow)ducts in females and the normalappearance of the ovary. (D) Wolffianduct (*) presents a similarly degeneratedaspect in males. E to H show details ofthe Wolffian ducts in wild-type andmutant females (E,F, respectively) andwild-type and mutant males (G,H,respectively). Degenerated ducts arefound inside the remnant of theurogenital ridge shown in Fig. 2. wd,Wolffian duct; md, Müllerian duct; k,kidney; vd, vas deferens.


Fig. 6. (A,B) Cross sections of femalesat sacral level of respectively wild-typeand mutant E17.0 embryos;(C,D) similar sections at a pelvic level.Observe that mutants lack uterinehorns, vagina and ureters. Arrowsmark the places of the missingstructures. r, rectum; ur, ureter; ut,uterine horn; b, bladder; v, vagina.

Fig. 7. Pax-2 expression in the developing Wolffian and Müllerianducts. (A,B) Phase contrast and fluorescent images, respectively, ofcross sections through the Wolffian and Müllerian ducts at E13.5showing antibody detection of Pax-2 protein expression in wild-typeembryos. wd, Wolffian duct; md, Müllerian duct. Fig. 8. Analysis of Pax-2 and Six-2 expression during mesonephric

development. (A) A vibratome cross section at E9.5 after whole-mount in situ hybridization with Pax-2 probe. The white arrowheadmarks the Wolffian duct and the black arrow, the nephrogenic cord.(B,C) Cross sections of wild-type and mutant embryos, respectively,at E9.5 after whole-mount in situ hybridization with a Six-2 probe.

followed the development of these ducts using histologicalanalysis. In addition, we used the c-ret gene as a marker, atyrosine kinase receptor expressed in the Wolffian duct andureters (Pachnis et al., 1993). At 9.5 days of embryonic devel-opment (E9.5), when most rostral parts of Wolffian duct arebeing generated, there are no differences between mutant andwild-type embryos in morphology and expression of c-ret (Fig.3). However, while the Wolffian duct has reached the cloacaand strongly expresses c-ret at E10.5 in Pax-2+/+ embryos, inPax-2−/− littermates the Wolffian ducts do not reach the cloacaand have lost c-ret expression (Fig. 3). A similar situation is

observed in E11.5 embryos where the Wolffian ducts do notextend beyond somite 25 in the mutants (Fig. 3).

In E12.5 embryos, the parts of the Wolffian duct initiallyformed are degenerating in discontinuous spherical epithelialbodies (Fig. 4). Müllerian ducts develop in parallel to Wolffianducts around E13.0. At E13.5 Müllerian ducts are found inwild-type embryos along the entire length of the urogenital


Fig. 9. Metanephric development in Pax-2 mutants. (A,B) Crosssections at a metanephric level of E12.5 wild-type and mutantembryos, respectively. wd, Wolffian duct; ur, ureter; mm,metanephric mesenchyme. Arrowhead marks the place of themissing Wolffian duct and asterisk marks metanephric mesenchymesin the mutants. (C,D) Kidney cross sections at the level where theureter stems from the calyx of, respectively, wild-type andheterozygous E17.0 embryos. c, calix; ur, ureter.

Fig. 10. Expression analysis of Six-2 and c-ret during metanephrosdevelopment in Pax-2 mutant embryos. Vibratome cross sections atthe metanephric level of E11.5 embryos hybridized with c-ret (A,B)or Six-2 probes (C,D). (A,C) Wild-type embryos; (B,D) mutantembryos. ur, ureter; w, Wolffian duct at the level where the ureterbuds out.

ridge parallel to the Wolffian ducts and reaching the cloaca. Atthe same stage, Müllerian ducts are only present at the uppermost levels of the genital ridge in Pax-2−/− embryos (data notshown). By E16.5 the part of the Müllerian duct initiallyformed has degenerated and the spherical bodies remainingfrom the degeneration of the Wolffian duct have lost theepithelial morphology in both sexes (Fig. 5). The degeneratedremains from both ducts are contained within the remnant ofthe urogenital ridges flanking the gonads (Fig. 5). Sections ata lower level at this stage show the absence of uterine hornsand vagina in mutant females (Fig. 6).

We have analysed the expression of Pax-2 protein in thedeveloping urogenital ridge at E13.5 when both Wolffian andMüllerian ducts are present. At this stage, Pax-2 is expressedin both developing Wolffian and Müllerian ducts, suggestingan autonomous function of Pax-2 in the development of bothducts (Fig. 7).

Development of mesonephros in Pax-2 mutantembryosThe mesonephros is a transient embryonic kidney consistingof a linear array of mesonephric tubules connecting to theWolffian duct (Saxén, 1987). Formation of the mesonephrictubules requires epithelial transformation of the nephrogeniccord mesenchyme flanking the Wolffian duct and is concomi-tant with Pax-2 expression (Dressler et al., 1990 and Fig. 8).Pax-2−/− embryos do not develop mesonephric tubules (Fig. 4),

therefore, it is likely that Pax-2 is needed for the epithelial tran-sition in the nephrogenic cord.

We have analyzed tubule formation using the homeoboxgene Six-2 as a marker (Oliver et al., 1995). While initiallyweakly expressed in the nephrogenic cord, Six-2 is up-regulated precluding tubule formation at about E10 (Fig. 8). Incontrast, Pax-2−/− embryos keep a low expression level of Six-2 at E10, not showing the up-regulation typical of this stage.Mesonephric tubules regress in the female but persist in themale where they form the ductuli efferentes, also missing inthe mutants (Fig. 2).

Metanephros development in Pax-2 mutantsDevelopment of the definitive kidney starts at about E11 whenthe ureter buds out from the posterior part of the Wolffian ductand contacts the metanephric mesenchyme, a morphologicallydistinct group of posterior nephrogenic cord cells (Grobstein,1953, 1955). Ultimately, the mesenchyme will undergo epithe-lial transformation to generate the definitive nephrons(Eckblom et al., 1981). Pax-2 is expressed in both the ureterand induced mesenchyme (Dressler et al., 1990). In Pax-2−/−

embryos, metanephric development does not take place sinceureter buds are absent (Fig. 9), and the ureter marker c-ret isnot expressed (Fig. 10). However, there is a morphologicallydistinct metanephric mesenchyme (Fig. 9). Parallelingmesonephric development, Six-2 is expressed weakly inuninduced mesenchyme and strongly up-regulated afterinvasion of the mesenchyme by the ureter bud at E11.5. Incontrast, mutant embryos still express Six-2 at similar levels tothe uninduced mesenchyme (Fig. 10).

The function of Pax-2 in metanephric development appears


so crucial that even one wild-type copy of the gene is notenough to provide normal kidney development. Pax-2+/−

animals frequently develop hypoplastic kidneys mainly due toreduced calyces and upper part of the ureters, suggestingdefects in ureter branching and/or cell proliferation, but also toa thinner cortical region where a reduced number of develop-ing nephrons is found, suggesting a decreased rate of epi-thelial transformation and/or cell proliferation (Fig. 9).

DISCUSSION

Pax-2 and the development of epithelial componentsof the intermediate mesodermThe defects found in urogenital system development in Pax- 2mutant mice point to a critical function during intermediatemesoderm differentiation. This mesoderm undergoes a seriesof sequential inductions that transform it into different epi-thelial components of the kidney and the urogenital tracts. Themesenchymal cells of the intermediate mesoderm are uniquein this respect, as most other epithelial tissues are formed bybranching morphogenesis of pre-existing epithelium. Pax-2 isexpressed during formation and differentiation of all epithelialstructures derived from intermediate mesoderm, consistentwith the developmental defects in Pax-2 mutants. In Pax-2mutants, we observe defects not only in the formation of theWolffian duct and ureter but also in tubule formation duringmesonephric development. The Wolffian duct appears normalat E9-10 but fails to extend caudally towards the cloaca.Rather, by E11, the duct begins to degenerate and appears asdiscontinuous epithelium with enlarged lumens. Although ductformation appears normal in the region of the mesonephros,Pax-2 mutant mice fail to form mesonephric tubules. Themutant Wolffian duct may not be competent to induce tubuleformation despite its normal morphology. However, it is likelythat the Pax-2 mutant mesonephric mesenchyme is unable toundergo epithelial transformation, as mesonephric tubuleformation can occur independent of the Wolffian duct(Etheridge, 1968; Croisille et al., 1976). Therefore, the defectsin mesonephric tubules observed in Pax-2−/− embryos mostprobably result from an autonomous function of Pax-2 intubule formation. Likewise, the defects observed in theWolffian duct formation and posterior degeneration are mostlikely due to an autonomous function of Pax-2 in the duct cells.However, we cannot exclude the possibility that the defects inthe Wolffian duct may result at least partially from an impairedinteraction with the mutant mesenchyme.

Because the ureter fails to form, it is not possible to studytubule formation during metanephric development in Pax-2mutant embryos. However, in vitro experiments using Pax-2antisense oligonucleotides (Rothenpieler and Dressler, 1993)have suggested that Pax-2 is also necessary for epithelial trans-formation in metanephric tubule formation. In our studies andthose by Keller et al. (1994), heterozygous mice carrying anormal Pax-2 allele show reduced kidney size and an attenu-ated nephrogenic zone. Thus, a Pax-2 function during prolif-eration and epithelial differentiation of the metanephric mes-enchyme is clearly indicated.

The function of Pax-2 is thus essential for the developmentof all epithelial components derived from the intermediatemesoderm, irrespective of their time of appearance ormechanism of induction. Supporting this view, the develop-

ment of the Müllerian duct is also impaired. The Müllerian ductand the Wolffian duct both originate from the intermediatemesoderm; however, each duct consists of unique cell typessince they react in opposite ways to sex hormones (Jost andMagre, 1984). Epithelial components are not the only deriva-tives from the intermediate mesoderm. Mesenchymal cellsfrom the mesonephros migrate during gonadal developmentand populate testis where they contribute to interstitial cellpopulations (Buehr et al., 1993). Quite remarkably, in Pax-2mutants, testis appear completely normal in structure and celltype composition (data not shown), supporting the view thatPax-2 function is specifically needed for development of theepithelial components of the intermediate mesoderm.

Among the genes required for urogenital development are c-ret, WT-1 and Wnt-4. WT-1 is expressed in the uninducedmetanephric mesenchyme and enables these cells to respond toinduction by the ureteric bud. In the absence of WT-1 functionin the mesenchyme, there is no ureteric bud outgrowth fromthe adjacent Wolffian duct. Thus, the outgrowth of the ductmay depend on positional cues originating from themetanephric mesenchyme. These cues may be received by thereceptor-type kinase c-ret, which is expressed in the Wolffianduct and subsequently in the tips of the growing ureter bud.Finally, in the absence of the secreted factor Wnt-4, tubuleformation does not take place in metanephros, despite normalureter growth and branching (Stark et al., 1994).

Because Pax-2 is a transcription factor, it may be requiredto regulate the expression of genes essential for the differen-tiation of the epithelial components of the urogenital system.We have observed alterations in the expression pattern of c-retand Six-2 during the development of the affected structures.However, it is likely that these alterations are the result ofimpaired development of the structures implicated, rather thanimplying direct control by Pax-2 of the expression of thesegenes. In support of this view, expression of the c-ret gene wasnormal in the Wolffian duct and altered only after the ductstarted degeneration.

Pax-2 and kidney diseaseThere is suggestive evidence that links human kidney diseaseswith altered Pax-2 expression. During normal kidney devel-opment, Pax-2 is repressed after tubule formation (Dressler etal., 1990). Overexpression of Pax-2 is incompatible withnormal kidney development in transgenic mice and results inkidney abnormalities similar to human nephrotic syndromes(Dressler et al., 1993). Persistent Pax-2 expression is found inWilm’s tumor (Dressler and Douglas, 1992; Eccles et al.,1992) and in adult renal carcinoma where it is required forcontinued proliferation (Gnarra and Dressler, 1995). Partialloss of function by reducing the Pax-2 gene dosage throughmutation of one of the alleles, either in humans or mice, resultsin hypoplastic kidneys (Keller et al., 1995; Sanyanusin et al.,1995; this report). Appropriate timing and dosage of Pax-2expression is therefore essential to accomplish proper devel-opment of the kidney. There are at least four haplo-insufficientPax genes, Pax-1, Pax-2, Pax-3 and Pax-6. Single gene haplo-insufficiency is rather rare; for the majority of genes one copyis enough to ensure wild-type function. The unusual highincidence of haplo-insufficient phenotypes among Pax genessuggests that, despite the heterogeneity in their developmentalfunctions, an underlying common molecular mechanism for


Pax protein function may exist that depends quantitatively onthe level of protein present in the cell.

Pax-2 is also expressed during central nervous system(CNS) development in eye, ear and midbrain (Nornes et al.,1990; Püschel et al., 1993). We have observed defects in thedevelopment of these three structures in Pax-2 mutant mice(unpublished data). Eye defects have also been reported in Krdmice and human PAX-2 mutants. Therefore, Pax-2 mutantmice provide an excellent animal model for the study of bothCNS and urogenital system human hereditary diseases.

We thank S. Mashur M. Breuer and R. Scholz for excellenttechnical assistance, G. Oliver, M. Kessel, F. Pituello, K. Chowdhuryand E. Stuart for critical comments on the manuscript, B. Sosa-Pinedafor advice on the cryostat sectioning, F. Constantini for the c-retprobe, G. Oliver for the Six-2 probe and R. Altschäffel for excellentphotographic work. M. T. received support from EMBO andEuropean Community Human Genome Analysis program.

REFERENCES

Balling, R., Deutsch, U. and Gruss, P. (1988). undulated, a mutation affectingthe development of the mouse skeleton has a point mutation in the paired boxof Pax-1. Cell 55, 531-535.

Bober, E., Franz, T., Arnold, H. H., Gruss, P. and Tremblay, P. (1994).Pax-3 is required for the development of limb muscles: a possible role for themigration of dermomyotomal muscle progenitor cells. Development 120,603-612.

Buehr, M., Gu, S. and McLaren, A. (1993). Mesonephric contribution totestis differentiation in the fetal mouse. Development 117, 273-281.

Chalepakis, G., Tremblay, P., and Gruss, P. (1992). Pax genes, mutants andmolecular function. J. Cell Sci. Supplement 16, 61-67.

Croisille, Y., Gumpel-Pinot, M. and Martin, C. (1976). Embryologieexpérimentale. La differenciation des tubes sécréteurs du rein chez lesoiseaux: effects des inducteurs hétérogènes. C. R. Acad. Sci. Paris Ser. 282,1987-1990.

Dressler, G. R., Deutsch, U., Chowdhury, K., Nornes, H. O. and Gruss, P.(1990). Pax-2, anew paired-box-containing gene and its expression in thedeveloping excretory system. Development 109, 787-795.

Dressler, G. R. and Douglas, E. C. (1992). Pax-2 is a DNA-binding proteinexpressed in embryonic kidney and Wilms tumor. Proc. Nat. Acad. Sci. USA89, 1179-1183.

Dressler, R., Wilkinson, J. E., Rothenpieler, U. W., Patterson, L. T.,Williams-Simons, L. and Westphal, H. (1993). Deregulation of Pax-2expression in transgenic mice generates kidney abnormalities. Nature 362,65-67.

Eccles, M. R., Wallis, L. J., Fidler, A. E., Spur, N. K., Goodfellow, P. J. andReeve, A. E. (1992). Expression of the Pax-2 gene in human fetal kidney andWilm’s tumor. Cell Growth Diff. 3, 279-289.

Eckblom, P., Miettinen, A., Virtanen, I., Wahlström, A. D. and Saxén, L.(1981). In Vitro segregation of the metanephric nephron. Dev. Biol. 84, 88-95.

Epstein, D. J., Vekemans, M. and Gros P. (1991). Splotch (Sp2h), a mutationaffecting the development of the mouse neural tube, shows a deletion withinthe paired homeodomain of Pax-3. Cell 67, 767-774.

Etheridge, A. L. (1968). Determination of the mesonephric kidney. J. Exp.Zool. 169, 357-370. (1968)

Gnarra, J. R. and Dressler, G. R. (1995). Expression of Pax-2 in renal cellcarcinoma and growth inhibition by antisense oligonucleotides. Cancer Res.(In Press.

Grobstein, C. (1953). Inductive epithelio-mesenchymal interaction in culturedorgan rudiments of the mouse Science 118, 52-55.

Grobstein, C. (1955). Tissue interactions in the morphogenesis of the mousemetanephros. J. Exp. Zool. 130, 319-340.

Gruss, P. and Walther, C. (1992). Pax in development. Cell 69, 719-722. Hadler, G., Callaerts, P. and Gehring, W. J. (1995). Induction of ectopic eyes

by targeted expression of the eyeless gene in Drosophila. Science 267, 1788-1792.

Hill, R. E., Favor, J., Hogan, B. L. M., Ton, C. C. T., Saunders, G. F.,Hanson, I. M., Prosser, J., Jordan, T., Hastie, N. D. and Van Heyningen,V. (1991). Mouse Small eye results from mutations in a paired-likehomeobox-containing gene. Nature 354, 522-525.

Jost, A. and Magre, S. (1984). Testicular development phases and dualhormonal control of sexual organogenesis. In Sexual Differentiation: Basicand clinical Aspects. (Ed. Serio et al.) pp 1-15. New York: Raven Press.

Keller, S. A., Jones, J. M., Boyle, A. Barrow, L. L., Killen, P. D., Green, D.G., Kapousta, N. V., Hitchcock, P. F., Swank, R. T. and Meisler, M. H.(1994). Kidney and retinal defects (Krd), a transgene-induced mutation witha deletion of mouse chromosome 19 that includes the Pax-2 locus. Genomics23, 309-320.

Kreidberg, J. A., Sariola, H., Loring, J. M., Maeda, M., Pelletier, J.,Housman, D. and Jaenisch, R. (1993). WT-1 is required for early kidneydevelopment. Cell 74, 679-691.

Nagy, A. and Rossant, J. (1993). Production of completely ES cell-derivedfetuses. In Gene Targeting, A Practical Approach. (Ed. A. Joyner) pp 147-179. Oxford: IRL Press.

Nornes, H. O., Dressler, G. R., Knapik, E. W., Deutsch, U. and Gruss, P.(1990). Spatially and temporally restricted expression of Pax-2 duringmurine neurogenesis. Development 109, 797-809.

Oliver, G., Wehr, R., Jenkins, N. A., Copeland, N. G., Cheyette, B. N. R.,Hartenstein, V., Zipursky, S. L. and Gruss, P. (1995). Homeobox genesand connective tissue patterning. Development 121, 693-705.

Pachnis, V., Mankoo, B. and Constantini, F. (1993). Expression of the c-retproto-oncogene during mouse embryogenesis. Development 119, 1005-1017.

Püschel, A. W., Westerfield, M. and Dressler, G. R. (1992). Comparativeanalysis of Pax-2 protein distributions during neurulation in mice andZebrafish. Mech. Dev. 38, 197-208.

Rothenpieler, U. W. and Dressler, G. R. (1993). Pax-2 is required formesenchyme-to-epithelium conversion during kidney development.Development 119, 711-720.

Sanyanusin, P., Schimenti, L. A., McNoe, L. A., Ward, T. A., Pierpont, M.E. M., Sullivan, M. J., Dobyns, W. B. and Eccles, M. R. (1995). Mutationof the Pax-2 gene in a family with optic nerve colobomas, renal anomaliesand vesiculoureteral reflux. Nature Genet. 9, 358-363.

Saxén, L. (1987). Organogenesis of the Kidney. Developmental and CellBiology Series 19 (ed. P. W. Barlow, P. B. Green and C. C. White)Cambridge: Cambridge University Press.

Schuchardt, A., D’Agati, V., Larsson-Blomberg, L., Constantini, F. andPachnis, V. (1994). Defects in the kidney and enteric nervous system of micelacking the tyrosine kinase receptor Ret. Nature 367, 380-383.

Stark, K., Vainio, S., Vassileva, G. and McMahon, A. P. (1994). Epithelialtransformation of metanephric mesenchyme in the developing kidneyregulated by Wnt-4. Nature 372, 679-683.

Strachan, T. and Read A. P. (1994). Pax genes. Curr. Op. Dev. Biol. 4, 427-438.

Urbanek, P.,Wang, Z., Fetka, I., Wagner, E. F. and Busslinger, M. (1994).Complete block of early B cell differentiation and altered patterning of theposterior midbrain in mice lacking Pax-5/BSAP. Cell 79, 901-912.

Wilkinson, D. G. (1992). Whole mount in situ hybridization of vertebrateembryos. In In situ hybridization, A Practical Approach. (ed. D. Wilkinson)pp 75-84. Oxford: IRL Press.

Wurst, W. and Joyner, A. L. (1993). Production of targeted embryonic stemcell clones. In Gene Targeting, A Practical Approach. (ed. A. Joyner) pp 33-62. Oxford: IRL Press.

(Accepted 12 September 1995)

pax-2controls multiple steps of urogenital development · the urogenital system is derived...

Documents